Risk for intimate partner violence : an investigation of the psychometric properties of the Spousal Assault Risk Assessment Guide in a New Zealand population : a thesis presented in partial fulfilment of the requirements for the degree of Doctorate of Clinical Psychology at Massey University, Wellington, New Zealand

International and national studies have consistently shown intimate partner violence is a common phenomenon that cuts across all societies, education and socioeconomic levels, and ethnic and cultural groups. The impact of which includes negative physical and mental health consequences for the victims. Risk assessments may play a role in assisting the management and/or prevention of harm. Assessment of an offender’s risk of future violence play a central role in decision making pertaining to that person’s sentencing, community release, case management, and public safety concerns. Yet the assessments also need to ensure that the rights of the individual being assessed are not violated by misclassification. One method for addressing this issue is to ensure that risk assessment measures are accurate, that is, the measure is reliable and valid. In New Zealand to date, no intimate partner violence risk assessment tools have been evaluated. The current study, therefore, aims to fill this void by investigating the reliability and validity of the Spousal Assault Risk Assessment (SARA) guide. This was achieved in three parts, using a sample of 43 men recruited from community based stopping violence programmes. Part One evaluated the internal consistency and interrater reliability of the SARA, Part Two evaluated the convergent and discriminant validities, and Part Three, which employed a prospective design with 36 participants from the total sample, evaluated the predictive validity and incremental validity of the dynamic risk factors. The findings indicated that while the internal consistency, and convergent, discriminant, and predictive validates were adequate, the dynamic risk factors did not evidence incremental validity over the static risk factors, and the interrater reliability was variable. In addition, it was found that the source of information provided to the observers impacted on the resulting agreement coefficients. Therefore, before the SARA is implemented as a risk assessment measure in New Zealand methods for improving the interrater reliability and exploration of the usefulness of the dynamic risk factors in reducing risk should be explored